Cantera Preconfig

Main article: Cantera

The following is my Cantera preconfig file, used to install Cantera version 1.8.0.

#!/bin/sh
#
#######################################################################
#
#           Cantera Configuration File
#
#  Edit this file to control how Cantera is built. Parameters can be set
#  here, or alternatively environment variables may be set before calling
#  this script.
#
#  The default configuration uses GNU compilers (gcc/g++/g77) and
#  builds as much of Cantera and its language interfaces as it can
#  (e.g. if MATLAB is installed on your system, the MATLAB toolbox
#  will be built automatically, otherwise it will be skipped. On linux
#  or Mac OS X, this default configuration should work, and most
#  likely you don't need to edit this file at all - just run it.
#  
#  NOTE: if you DO make changes to this file, save it with another name
#  so that it will not be overwritten if you update the source
#  distribution. 

#######################################################################
#
# Set CANTERA_CONFIG_PREFIX to the directory where you want Cantera
# installed into. The default is /usr/local/Cantera
#
# If you define this to be <prefix>, then instead of running this
# script as ./preconfig --prefix=<prefix> you can just run it as
# ./preconfig.

CANTERA_CONFIG_PREFIX=${CANTERA_CONFIG_PREFIX:="/Users/charles/pkg/cantera-1.8.0"}

#----------------------------------------------------------------------
#         Language Interfaces
#----------------------------------------------------------------------
#
# Cantera has several programming language interfaces. Select the ones
# you want to build. The default is to try to build all language
# interfaces.
#
#
#----------------- Python --------------------------------------------
# 
# In addition to being one of the supported language interfaces,
# Python is used internally by Cantera, both in the build process and
# at run time (to process .cti input files). Therefore, you generally
# need to have Python on your system; if you don't, first install it
# from http://www.python.org before proceeding with the installation
# of Cantera.
#
# If you plan to work in Python, or you want to use the graphical
# MixMaster application, then you need the full Cantera Python
# Package. If, on the other hand, you will only use Cantera from some
# other language (e.g. MATLAB or Fortran 90/95) and only need Python
# to process .cti files, then you only need a minimal subset of the
# package (actually, only one file).

# Set PYTHON_PACKAGE to one of these four strings:
#    full      install everything needed to use Cantera from Python
#    minimal   install only enough to process .cti files
#    none      Don't install  or run any Python scripts during the 
#              build process
#    default   try to do a full installation, but fall back to a minimal
#              one in case of errors

PYTHON_PACKAGE=${PYTHON_PACKAGE:="full"}

# Cantera needs to know where to find the Python interpreter.  If
# PYTHON_CMD is set to "default", then the configuration process will
# look for the Python Interpreter somewhere on your PATH, either as
# 'python2' or as 'python'.  If it is not on the PATH, or has a
# different name, set this to the full path to the Python interpreter.
PYTHON_CMD=${PYTHON_CMD:="/Library/Frameworks/Python.framework/Versions/2.6/bin/python"}


# The Cantera Python interface can be built with either the numarray
# numeric, or numpy packages. Set this to "y" to use Numeric.
USE_NUMERIC=${USE_NUMERIC:="n"}
#USE_NUMARRAY=${USE_NUMARRAY:="n"}

# The Cantera Python interface can be built with either the numarray
# or Numeric packages. Set this to "y" to use numpy, or 'n'
# to use numarray. set USE_NUMERIC to 'n' also if you want to use numarray.
# Using numpy is preferred, and is the supported option going forwards.
USE_NUMPY=${USE_NUMPY:="y"}

# If numpy was installed using the --home option, set this to the
# home directory for numpy. This will be needed for all numpy installations
# that don't put the include files into python's native include directory.

# NOTE: modified because Cantera tacks "/include/python" onto the end of NUMPY_HOME, 
# which means the compiler doesn't point to include files in the correct directory
NUMPY_INC_DIR=${NUMPY_INC_DIR:="/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/numpy/core/include/numpy"}
export NUMPY_INC_DIR

# If numarray was installed using the --home option, set this to the
# home directory for numarray.
#NUMARRAY_HOME=${NUMARRAY_HOME:="/Library/Frameworks/Python.framework/Versions/2.6/lib/python2.6/site-packages/numpy/numarray"}


# If you want to install the Cantera Python package somewhere other
# than the default 'site-packages' directory within the Python library
# directory, then set this to the desired directory. This is useful when
# you do not have write access to the Python library directory. 
#CANTERA_PYTHON_HOME=${CANTERA_PYTHON_HOME:="$HOME/python_modules"}
  
# Set this to 'y' when site packages must be put in system directories
# but Cantera tutorials must be put in user space.  An alternative to
# doing this is to keep this setting on "n" . Then everything is 
# put into the user space by running this script
# with the 'prefix' option. The environment variable PYTHONPATH
# may then be set to tell Python where to find the Cantera package.
#
SET_PYTHON_SITE_PACKAGE_TOPDIR=${SET_PYTHON_SITE_PACKAGE_TOPDIR:="n"}

# If the SET_PYTHON_SITE_PACKAGE_TOPDIR is set to y, the entry below
# sets the location where the Cantera site package is installed.
#
PYTHON_SITE_PACKAGE_TOPDIR=${PYTHON_SITE_PACKAGE_TOPDIR:="/usr/local"} 

#----------- MATLAB --------------------------------------------------

# This variable controls whether the MATLAB toolbox will be built. It
# should be set to one of these strings:
#
#      "default"     the 'configure' script will try to build a simple 
#                    MATLAB MEX file. If this succeeds, BUILD_MATLAB_TOOLBOX
#                    will be set to "y", otherwise it will be set to "n".
#                    
#      "y"           build the MATLAB toolbox.
#      "n"           do not build the MATLAB toolbox, even if MATLAB is 
#                    installed.
#
# Note that you may need to run 'mex -setup' within MATLAB to
# configure it for your C++ compiler before running this script.
BUILD_MATLAB_TOOLBOX=${BUILD_MATLAB_TOOLBOX:="n"}
                             
#----------- Fortran 90/95 --------------------------------------------------

# This variable controls whether the Fortran 90/95 interface will be
# built. It should be set to one of these strings:
#
#      "default"    The 'configure' script will attempt to compile a small
#                   Fortran 90/05 program; if this succeeds, 
#                   BUILD_F90_INTERFACE will be set to "y", and otherwise to "n".
#      "y"          Build the Fortran 90/95 interface.
#      "n"          Do not build the Fortran 90/95 interface, even if a 
#                   Fortran 90/95 compiler is installed.

BUILD_F90_INTERFACE=${BUILD_F90_INTERFACE:="no"}

# The Fortran 90/95 compiler. If set to "default", the script will
# look for a Fortran 90/95 compiler on your system by the name of
# "f95", "gfortran", or "g95".
F90=${F90:="gfortran"}

# Compiler option flags for the Fortran 90/95 compiler. If you are
# using the Absoft or the NAG compiler, additional options specific to
# these compilers will be added automatically, and you do not need to
# specify them here.  Otherwise, add any required compiler-specific
# flags here.
F90FLAGS=${F90FLAGS:='-O3'}

#---------------------------------------------------------------------
#          Purify
#---------------------------------------------------------------------

#  Add entries here for compiling Cantera with Purify extensions
#  This command is put in front of every compilation and linking
#  step
PURIFY=${PURIFY:=""}

#----------------------------------------------------------------------
#         Customizations / Extensions
#----------------------------------------------------------------------
#
# You can build your own libraries as part of the Cantera build process.
# This allows you to derive your own classes from those provided by 
# Cantera and build them automatically along with the rest of Cantera.
# All you need to do is specify the directory where your source code is
# located. This capability is not yet fully functional, but should work
# for C++ applications.
USER_SRC_DIR="Cantera/user"  # don't change this

# In some cases, extra printing has been put into the code to aid
# in debugging and verification. In order for the printing to occur
# an additional switch must be turned on. However, compilation
# of the printing is affected by the following DEBUG_MODE option.
# This is turned off here by default. 
DEBUG_MODE=${DEBUG_MODE:='n'}

#----------------------------------------------------------------------
#         Kernel Configuration
#----------------------------------------------------------------------

# If you are only planning to use a portion of Cantera, you may only
# need a stripped-down kernel. Set those features you want enabled to
# 'y', and set those you want to skip to 'n' (actually anything but
# 'y') or comment them out. Some features are dependent on others; for
# example, enabling 'CK' automatically enables KINETICS and THERMO.

# If you only need to use Cantera to evaluate thermodynamic, kinetic,
# and transport properties, it is sufficient to enable only KINETICS
# and TRANSPORT.

# Note: if you are building the full Python interface or the MATLAB
# interface, it is necessary to build the full kernel.

# thermodynamic properties 
ENABLE_THERMO='y'

######################################################################
# optional phase types. These may not be needed by all users. Set them
# to 'n' to omit them from the kernel.

WITH_LATTICE_SOLID=${WITH_LATTICE_SOLID:="y"}
WITH_METAL=${WITH_METAL:="y"}
WITH_STOICH_SUBSTANCE=${WITH_STOICH_SUBSTANCE:="y"}
WITH_SEMICONDUCTOR=${WITH_SEMICONDUCTOR:="y"}
WITH_ADSORBATE=${WITH_ADSORBATE:="y"}
WITH_SPECTRA=${WITH_SPECTRA:="y"}

# This flag enables the inclusion of accurate liquid/vapor equations
# of state for several fluids, including water, nitrogen, hydrogen,
# oxygen, methane, andd HFC-134a.
WITH_PURE_FLUIDS='y'

# This flag enables the inclusion of ideal solution capabilities
WITH_IDEAL_SOLUTIONS=${WITH_IDEAL_SOLUTIONS:="y"}

# Enable expanded electrochemistry capabilities, include thermo
# models for electrolyte solutions
WITH_ELECTROLYTES=${WITH_ELECTROLYTES:="y"}

# Enable generating phase models from PrIMe models. For more
# information about PrIME, see http://www.primekinetics.org
# WARNING: Support for PrIMe is experimental!
WITH_PRIME=${WITH_PRIME:="y"}

# Enable changing the 298K heats of formation directly
# via the C++ layer. To do this, modify the variable below
# to have a value of "y"
WITH_H298MODIFY_CAPABILITY=${WITH_H298MODIFY_CAPABILITY:="n"}

######################################################################
# if set to 'y', the ck2cti program that converts Chemkin input files
# to Cantera format will be built. If you don't use Chemkin format
# files, or if you run ck2cti on some other machine, you can set this to 'n'. 
ENABLE_CK=${ENABLE_CK:='y'}

######################################################################

# Enable homogeneous kinetics
WITH_KINETICS='y'

# Enable heterogeneous kinetics (surface chemistry). This also enables
# charge transfer reactions for electrochemistry.
WITH_HETERO_KINETICS='y'

# Enable reaction path analysis
WITH_REACTION_PATHS='y'

# Enable vcs equilibrium package for nonideal phases
WITH_VCSNONIDEAL=${WITH_VCSNONIDEAL:="n"}

######################################################################
# transport properties
ENABLE_TRANSPORT='y'  

# chemical equilibrium
ENABLE_EQUIL='y'      
     
# stirred reactor models
ENABLE_REACTORS='y'     

# One-dimensional flows
ENABLE_FLOW1D='y'

# ODE integrators and DAE solvers
ENABLE_SOLVERS='y'

# reaction path analysis
ENABLE_RXNPATH='y'         

# two phase pure fluids
ENABLE_TPX='y'

# write HTMl log files. Some multiphase equilibrium procedures can
# write copious diagnostic log messages. Set this to anything but 'y'
# to remove this capability (results in slightly faster equilibrium 
# calculations)
WITH_HTML_LOG_FILES='y'

#-----------------------------------------------------------------
#         CVODE / CVODES
#-----------------------------------------------------------------
# 
# Cantera uses the CVODE or CVODES ODE integrator to time-integrate
# reactor network ODE's and for various other purposes. An older
# version of CVODE comes with Cantera, but it is possible to use the
# latest version as well, which now supports sensitivity analysis
# (CVODES).  CVODES is a part of the 'sundials' package from Lawrence
# Livermore National Laboratory. Sundials is not distributed with
# Cantera, but it is free software that may be downloaded and
# installed separately. If you leave USE_SUNDIALS = 'default', then it
# will be used if you have it, and if not the older CVODE will be
# used. Or set USE_SUNDIALS to 'y' or 'n' to force using it or not.
# Note that sensitivity analysis with Cantera requires use of
# sundials.
#
# See: http://www.llnl.gov/CASC/sundials
#
USE_SUNDIALS=${USE_SUNDIALS:='y'}

# This is where you installed sundials if you used the --prefix option
# when you configured sundials. If you didn't use the prefix option,
# then comment this line out.
SUNDIALS_HOME=${SUNDIALS_HOME:="/Users/charles/pkg/sundials-2.3.0"}


# It is recommended that you install the newest release of sundials 
# (currently 2.3.0) before building Cantera. But if you want to use an 
# older version, set SUNDIALS_VERSION to the version you have.
# Acceptable values are '2.2' and '2.3' only; anything else will cause
# Cantera to not use sundials.
SUNDIALS_VERSION=${SUNDIALS_VERSION:='2.3'}

#----------------------------------------------------------------- 
#         BLAS and LAPACK
#-----------------------------------------------------------------
#
# Cantera comes with Fortran (or c) versions of those parts of BLAS and
# LAPACK it requires. But performance may be better if you use a
# version of these libraries optimized for your machine hardware. If
# you want to use your own libraries, set BLAS_LAPACK_LIBS to the
# string that should be passed to the linker to link to these
# libraries, and set BLAS_LAPACK_DIR to the directory where these
# libraries are located. Otherwise, leave these lines commented out.
#
#BLAS_LAPACK_LIBS='-llapack -lblas'
#BLAS_LAPACK_LIBS'-llapack -lf77blas -lcblas -latlas'
#BLAS_LAPACK_DIR='/usr/lib'
#
# Note that on Mac OSX, BLAS and LAPACK from the built-in 'Accelerate'
# framework are used by default, and nothing needs to be specified
# here.
#
# The options below do not need to be set if you are using the default
# libraries.
#
# Set to 'lower' or 'upper', depending on whether the procedure names
# in the libraries are lowercase or uppercase. If you don't know, run
# 'nm' on the library file (e.g. 'nm libblas.a') Note that the these options
# apply to both the BLAS and LAPACK libraries. 
LAPACK_NAMES='lower'
LAPACK_FTN_TRAILING_UNDERSCORE='y'

# Currently this must be set to 'y'.
LAPACK_FTN_STRING_LEN_AT_END='y'


#------------------------------------------------------------------
#       C++ compiler options
#------------------------------------------------------------------

# The number of bits to compile with, if different than the
# default on the computer
BITCOMPILE=${BITCOMPILE:=""}

# the C++ compiler to use.
CXX=${CXX:=g++}

# the C compiler to use. This is only used to compile CVODE and
# the Python extension module.
CC=${CC:=gcc}

# C++ compiler flags
CXXFLAGS=${CXXFLAGS:="-Wall"}

# the C++ flags required for linking. Uncomment if additional flags
# need to be passed to the linker.

# Ending libraries to tack onto the linking of all C++ programs
LCXX_END_LIBS=${LCXX_END_LIBS:="-lm"}

# The compiler flag to use to compile code that will be inserted into
# shared libraries.
PIC=${PIC:=""}

# the compiler option to create a shared library from object files
SHARED=${SHARED:="-dynamic"}
 
#-------------------------------------------------------------------
#      Thread Safety
#-------------------------------------------------------------------

# Cantera can be built so that it is thread safe. Doing so requires
# using procedures from the boost library, so if you want thread
# safety then you need to get and install boost (http://www.boost.org) 
# if you don't have it.  This is turned off by default, in which case
# boost is not required to build Cantera.

BUILD_THREAD_SAFE=${BUILD_THREAD_SAFE:="n"}

# where boost header and library files may be found
BOOST_INC_DIR=${BOOST_INC_DIR:="/Users/charles/pkg/boost_1_44_0"}
BOOST_LIB_DIR=${BOOST_LIB_DIR:="/Users/charles/pkg/boost_1_44_0/libs"}

# the boost thread library
BOOST_THREAD_LIB=${BOOST_THREAD_LIB:="boost_thread-mt-1_44"}

#-------------------------------------------------------------------
#      External Procedures
#-------------------------------------------------------------------

# Cantera uses several external software packages, which are all in
# the 'ext' directory.. These options control how these packages are
# built.

# For external procedures written in Fortran 77, both the original F77
# source code and C souce code generated by the 'f2c' program are
# included.  Set this to "y" if you want to build Cantera using the
# f2c-generated C sources instead of using the F77 sources in the ext
# directory.

# The default is to use the f2c code.  Set this to "n" to never do
# this on any platform.  Note, if set to "y", Cantera does not need a
# fortran compiler, and all fortran compiler parameters below are
# irrelevant.
#
BUILD_WITH_F2C=${BUILD_WITH_F2C:="y"}

# if you are building the external Fortran 77 procedures from the Fortran 
# source code, enter the compiler here. 
F77=${F77:=g77}

# Fortran 77 compiler flags. Note that the Fortran compiler flags must be set
# to produce object code compatible with the C/C++ compiler you are using.
FFLAGS=${FFLAGS:='-O3'}

# the additional Fortran flags required for linking, if any. Leave commented
# out if no additional flags are required.
LFORT_FLAGS="-L/usr/local/lib"

#------------------------------------------------------
#        Other Programs
#------------------------------------------------------

# The command to create a static library.
ARCHIVE=${ARCHIVE:="ar ruv"}   # linux
# ARCHIVE=${ARCHIVE:="libtool -static -o"}  # Mac OSX (this is set automatically)
# ARCHIVE=${ARCHIVE:="CC -xar -o"} # Solaris using SUNspro compilers

# The command to run 'ranlib' if it is needed. 
RANLIB=${RANLIB:="ranlib"}

# The command to run the 'make' utility. The Cantera Makefiles are
# compatible with the GNU make utility, so if your make utility
# doesn't work, try GNU make.
MAKE=${MAKE:=make}

# The command to run when "installing". This defaults now to the
# install-sh bourne shell script that comes with the autoconf package.
# It can be overridden here.
INSTALL_BIN=${INSTALL_BIN:=config/install-sh}

# The directory location of the graphviz program, dot.  dot is used
# for creating the documentation, and for making reaction path
# diagrams.  if "dot" is in your path, you can leave this unspecified.
#
# NOTE: MATLAB comes with a stripped-down version of 'dot'. If 'dot'
# is on your path, make sure it is not the MATLAB version!
#
GRAPHVIZDIR=${GRAPHVIZDIR:="/usr/local/bin/dot"}

# File extensions
CXX_EXT=${CXX_EXT:=cpp}
F77_EXT=${F77_EXT:=f}
F90_EXT=${F90_EXT:=f90}
#
# Default extension for executable binaries. The default for unix systems
# is the empty field. For windows systesms and cygwin, you should put
# ".exe" .
#
EXE_EXT=${EXE_EXT:=}


CT_SHARED_LIB=${CT_SHARED_LIB:=clib}

# The font to use in reaction path diagrams. This must be a font name
# recognized by the 'dot' program. On linux systems, this should be
# lowercase 'helvetica'.
RPFONT=${RPFONT:="Helvetica"}

# Don't change this.
CANTERA_VERSION=${CANTERA_VERSION:="1.8.0"}